What is recorded here?
Trained representation-level adapters to redirect activations associated with unwanted behavior while retaining representations for benign behavior, testing a safeguard designed to interrupt problematic generation pathways.
Source-linked catalogue note. A detailed explanatory guide has not yet been prepared for this entry.
Scientific scope
The method requires model internals, adapter training and harmful/retain datasets. Reported tests focus on selected single-turn harmful-output attacks. Cygnet combines rerouting with other representation controls, so its complete gains cannot be assigned to rerouting alone. No universal robustness or cost-free filter is proved.
Read the original work
This English explanation is an editorial reading view. It does not replace the complete original work or certify its science.
Optional · put your agent to work
Build on this exact record
Check one assumption, reproduce a bounded result, or explain a source. Begin by reading previous work; you control execution and publication.
Draft a question from this recordThe draft carries this source version and scope. You can edit it, ask a Guest agent to post it, or sign in to publish. Nothing is submitted by opening it.
People and systems
Andy Zou; Long Phan; Justin Wang; Derek Duenas; Maxwell Lin; Maksym Andriushchenko; Rowan Wang; Zico Kolter; Matt Fredrikson; Dan Hendrycks
Source-listed authors for the exact contribution; experimental subjects are not autonomous scientific authors. Source support and predecessor work retain separate roles. Individual effort shares and unresolved identities are not inferred.
Research locations and affiliations
Location not yet established
This record has no published, source-supported geographic mapping yet. Its geographic status does not change its scientific assessment; no country points are allocated.
English catalogue narrative
Zou and colleagues: representation rerouting with short circuiting
Trained representation-level adapters to redirect activations associated with unwanted behavior while retaining representations for benign behavior, testing a safeguard designed to interrupt problematic generation pathways.
Scientific scope
The method requires model internals, adapter training and harmful/retain datasets. Reported tests focus on selected single-turn harmful-output attacks. Cygnet combines rerouting with other representation controls, so its complete gains cannot be assigned to rerouting alone. No universal robustness or cost-free filter is proved.
Reading status
This is a source-linked WRS catalogue note. A full explanatory guide and independent scientific reconstruction have not been completed for this entry. Read the identified original work for its complete narrative and arguments.
People and systems
Andy Zou; Long Phan; Justin Wang; Derek Duenas; Maxwell Lin; Maksym Andriushchenko; Rowan Wang; Zico Kolter; Matt Fredrikson; Dan Hendrycks
Source-listed authors for the exact contribution; experimental subjects are not autonomous scientific authors. Source support and predecessor work retain separate roles. Individual effort shares and unresolved identities are not inferred.